* Reproduction Materials for "Playing the Women Card" - Study 1 
* Created by Erin Cassese

**** set working directory ****

set scheme s2mono

use study1.dta


* Table 1 - Sample Characteristics 

tab female
summ age
centile income 
tab edu
tab white
tab black
tab hispanic 
tab race_other
tab pid3
tab news_freq
tab reg_vote

* Figure 1 - Hostile and Benevolent Sexism by Race and Party 

graph bar (mean) hostile benevolent, over(female) over(pid3) ylabel(-.6(.2).6) ytitle("Mean of Hostile/Benevolent Sexism") title("A. Study 1")

* Table 2 - Effect of the Woman-Card Attack on Candidate Evaluations

regress clinton_therm hostile benevolent female age independent republican inc edu black hispanic race_other news_freq reg_vote if exposed==0, robust 
stdBeta, se
regress clinton_therm hostile benevolent female age independent republican inc edu black hispanic race_other news_freq reg_vote if exposed==1, robust
stdBeta, se
regress trump_therm hostile benevolent female age independent republican inc edu black hispanic race_other news_freq reg_vote if exposed==0, robust 
stdBeta, se
regress trump_therm hostile benevolent female age independent republican inc edu black hispanic race_other news_freq reg_vote if exposed==1, robust 
stdBeta, se
logit clinton_vote hostile benevolent female age independent republican inc edu black hispanic race_other news_freq reg_vote if exposed==0, robust 
stdBeta, se
logit clinton_vote hostile benevolent   female   age independent republican  inc edu black hispanic race_other  news_freq reg_vote if exposed==1, robust 
stdBeta, se

*** figure 2 *** 

summ hostile benevolent


regress clinton_therm hostile benevolent female age independent republican inc edu black hispanic race_other news_freq reg_vote if exposed==1, robust beta
margins, at(hostile=-1.214905) atmeans
margins, at(hostile=0) atmeans
margins, at(hostile=2.264225) atmeans
margins, at(benevolent=-1.473682) atmeans
margins, at(benevolent=0) atmeans
margins, at(benevolent=1.84105) atmeans

* Middle row of figures
    
regress trump_therm hostile benevolent female age independent republican inc edu black hispanic race_other news_freq reg_vote if exposed==1, robust beta
margins, at(hostile=-1.214905) atmeans
margins, at(hostile=0) atmeans
margins, at(hostile=2.264225) atmeans
margins, at(benevolent=-1.473682) atmeans
margins, at(benevolent=0) atmeans
margins, at(benevolent=1.84105) atmeans

* Bottom row of figures

logit clinton_vote hostile benevolent female age independent republican inc edu black hispanic race_other news_freq reg_vote if exposed==1, robust 
margins, at(hostile=-1.214905) atmeans
margins, at(hostile=0) atmeans
margins, at(hostile=2.264225) atmeans
margins, at(benevolent=-1.473682) atmeans
margins, at(benevolent=0) atmeans
margins, at(benevolent=1.84105) atmeans



clear 

insheet using "F2A.csv"
graph set window fontface "Times New Roman"


# delimit ;


twoway 


rcap cilo cihi order if lab==1,  color(black) ||


 
 scatter mean order if lab==1,   color(black) msize(medium) ||,

  			//addplot(pci 0.1185567 6 0.1185567 10) 
			title("" , color(black))
  			xtitle("Hostile Sexism")

  			xlab(
	
			16 "Max"
			9 "Mean" 
			2 "Min" 
			
  			, nogrid  labsize(medium))
  			ylab(, nogrid  labsize(medium)) 
  			ytitle("Predicted Clinton Therm.", size(medium))
  			xsize(6)
			legend(off)
			legend(size(small))
			ylab(20(10)60)
            graphregion(fcolor(white) color(white) lcolor(white))
 			plotregion(fcolor(white) color(white) lcolor(white));
//make legend small 
 		
#delimit cr

clear

insheet using "F2B.csv"
graph set window fontface "Times New Roman"


# delimit ;


twoway 


rcap cilo cihi order if lab==1,  color(black) ||


 
 scatter mean order if lab==1,   color(black) msize(medium) ||,

  			//addplot(pci 0.1185567 6 0.1185567 10) 
			title("" , color(black))
  			xtitle("Benevolent Sexism")

  			xlab(
	
			16 "Max"
			9 "Mean" 
			2 "Min" 
			
  			, nogrid  labsize(medium))
  			ylab(, nogrid  labsize(medium)) 
  			ytitle("Predicted Clinton Therm.", size(medium))
  			xsize(6)
			legend(off)
			legend(size(small))
			ylab(20(10)60)
            graphregion(fcolor(white) color(white) lcolor(white))
 			plotregion(fcolor(white) color(white) lcolor(white));
//make legend small 
 		
#delimit cr

clear

insheet using "F2C.csv"
graph set window fontface "Times New Roman"


# delimit ;


twoway 


rcap cilo cihi order if lab==1,  color(black) ||


 
 scatter mean order if lab==1,   color(black) msize(medium) ||,

  			//addplot(pci 0.1185567 6 0.1185567 10) 
			title("" , color(black))
  			xtitle("Hostile Sexism")

  			xlab(
	
			16 "Max"
			9 "Mean" 
			2 "Min" 
			
  			, nogrid  labsize(medium))
  			ylab(, nogrid  labsize(medium)) 
  			ytitle("Predicted Trump Therm.", size(medium))
  			xsize(6)
			legend(off)
			legend(size(small))
			ylab(20(10)60)
            graphregion(fcolor(white) color(white) lcolor(white))
 			plotregion(fcolor(white) color(white) lcolor(white));
//make legend small 
 		
#delimit cr

clear

insheet using "F2D.csv"
graph set window fontface "Times New Roman"


# delimit ;


twoway 


rcap cilo cihi order if lab==1,  color(black) ||


 
 scatter mean order if lab==1,   color(black) msize(medium) ||,

  			//addplot(pci 0.1185567 6 0.1185567 10) 
			title("" , color(black))
  			xtitle("Benevolent Sexism")

  			xlab(
	
			16 "Max"
			9 "Mean" 
			2 "Min" 
			
  			, nogrid  labsize(medium))
  			ylab(, nogrid  labsize(medium)) 
  			ytitle("Predicted Trump Therm.", size(medium))
  			xsize(6)
			legend(off)
			legend(size(small))
			ylab(20(10)60)
            graphregion(fcolor(white) color(white) lcolor(white))
 			plotregion(fcolor(white) color(white) lcolor(white));
//make legend small 
 		
#delimit cr

clear

insheet using "F2E.csv"
graph set window fontface "Times New Roman"


# delimit ;


twoway 


rcap cilo cihi order if lab==1,  color(black) ||


 
 scatter mean order if lab==1,   color(black) msize(medium) ||,

  			//addplot(pci 0.1185567 6 0.1185567 10) 
			title("" , color(black))
  			xtitle("Hostile Sexism")

  			xlab(
	
			16 "Max"
			9 "Mean" 
			2 "Min" 
			
  			, nogrid  labsize(medium))
  			ylab(, nogrid  labsize(medium)) 
  			ytitle("Pr(Clinton Vote)", size(medium))
  			xsize(6)
			legend(off)
			legend(size(small))
			ylab(.10(.10).60)
            graphregion(fcolor(white) color(white) lcolor(white))
 			plotregion(fcolor(white) color(white) lcolor(white));
//make legend small 
 		
#delimit cr

clear

insheet using "F2F.csv"
graph set window fontface "Times New Roman"


# delimit ;


twoway 


rcap cilo cihi order if lab==1,  color(black) ||


 
 scatter mean order if lab==1,   color(black) msize(medium) ||,

  			//addplot(pci 0.1185567 6 0.1185567 10) 
			title("" , color(black))
  			xtitle("Benevolent Sexism")

  			xlab(
	
			16 "Max"
			9 "Mean" 
			2 "Min" 
			
  			, nogrid  labsize(medium))
  			ylab(, nogrid  labsize(medium)) 
  			ytitle("Pr(Clinton Vote)", size(medium))
  			xsize(6)
			legend(off)
			legend(size(small))
			ylab(.10(.10).60)
            graphregion(fcolor(white) color(white) lcolor(white))
 			plotregion(fcolor(white) color(white) lcolor(white));
//make legend small 
 		
#delimit cr

clear


***** study 2 analysis ***** 


********************************************************************************
* Study 2 Analysis *************************************************************
********************************************************************************

use study2.dta

* Table 1 - Sample Characteristics 

tab female
summ age
centile income 
tab education
tab white
tab black
tab hispanic
tab otherrace
tab pid3
tab news_freq
tab registered

* Figure 1 - Hostile and Benevolent Sexism by Race and Party 

graph bar (mean) hostile benevolent, over(female) over(pid3) ylabel(-.6(.2).6) ytitle("Mean of Hostile/Benevolent Sexism") title("B. Study 2")

* Manipulation Checks

ttest tone, by(treatment) unequal
ttest anger, by(treatment) unequal
ttest enthusiasm, by(treatment) unequal
ttest anxiety, by(treatment) unequal

* Table 3 - Standardized 

regress anger treatment hostile benevolent treatment_hsexism treatment_bsexism female age independent republican income edu black hispanic other, robust 
stdBeta, se
regress enthusiasm treatment hostile benevolent  treatment_hsexism treatment_bsexism female age independent republican  income edu black hispanic other, robust 
stdBeta, se
regress anxiety treatment hostile benevolent  treatment_hsexism treatment_bsexism female age independent republican  income edu black hispanic other, robust 
stdBeta, se
regress participation treatment hostile benevolent  treatment_hsexism treatment_bsexism female age independent republican income edu black hispanic other anger anxiety enthusiasm, robust 
stdBeta, se

* Figure 3 

* Top Row
regress anger i.treatment##c.hostile benevolent  treatment_bsexism female age independent republican income edu black hispanic other, robust
margins treatment, at(hostile==-1.21) at (hostile==1.92)
marginsplot, name (gg1) ytitle("Predicted Anger") title("Anger and Hostile Sexism")
regress anxiety i.treatment##c.benevolent hostile  treatment_hsexism female age independent republican income edu black hispanic other, robust
margins treatment, at(benevolent==-1.36) at (benevolent==1.76)
marginsplot, name (gg2)  ytitle("Predicted Anxiety") title("Anxiety and Benevolent Sexism")
graph combine gg1 gg2, title("A. Emotional Reactions to the 'Woman Card' Attack")
* Bottom Row
eststo: regress participation i.treatment##c.hostile benevolent treatment_bsexism female age independent republican income edu black hispanic other anger anxiety enthusiasm, robust
margins treatment, at(hostile==-1.21) at (hostile==1.92)
marginsplot, name(a1, replace) level(84)  ytitle("Predicted Intent to Participate") title("Hostile Sexism")
eststo: regress participation i.treatment##c.benevolent hostile treatment_hsexism female age independent republican income edu black hispanic other anger anxiety enthusiasm, robust
margins treatment, at(benevolent==-1.36) at (benevolent==1.76)
marginsplot, name (a2, replace) level(84) ytitle("Predicted Intent to Participate") title("Benevolent Sexism")
graph combine a1 a2, title("B. Behavioral Reactions to the 'Woman Card' Attack")

* Vote Choice (p70)

xtile bsexism_3 = benevolent, nq(3)
xtile hsexism_3 = hostile, nq(3) 
tab presvote hsexism_3, col 
tab presvote bsexism_3, col





























